Rhodium-catalyzed asymmetric hydroformylation with taddol-based indolphos ligands

Article abstract of DOI:10.1021/om100236v

A small library of Taddol-based IndolPhos ligands 2a-g and their use in asymmetric hydroformylation (AHF) reactions are reported. Moderate to good enantioselectivities are obtained for styrene, vinyl acetate, and allyl cyanide up to 72%, 74%, and 63% ee, respectively. High b/l ratios are obtained, which results in a high net yield of the desired chiral aldehyde. An unprecedented temperature-dependent reversal of enantioselectivity is found when using ligands 2d,e, which are based on a xylyl-derived Taddol in the Rh-catalyzed AHF of styrene. Furthermore, these ligands give the opposite enantiomer of the product for vinyl acetate and allyl cyanide when compared to the other library members, all of which are based on (R,R)-tartaric acid. Ligands 2a and 2d display a similar kinetic profile, which is best described by type I kinetics. Deuterioformylation experiments have shown that insertion of the alkene into the Rh-H bond is irreversible under the conditions applied. High-pressure NMR studies indicate that the hydridobiscarbonyl rhodium species, which is the resting state of the catalyst, features equatorial-apical coordination of the ligands with a preference for the phosphine on the apical position, isomer A. However, in the case of ligand 2d the isomer in which the phosphoramidite occupies the apical position, B, is only marginally higher in free energy than isomer A. It is proposed that formation of the product takes place via isomer B for ligands 2d,e and via isomer A for all other ligands, explaining the observed reversal of enantioselectivity.

Full text of DOI:10.1021/om100236v

Organometallics 2010, 29, 2767–2776 2767
DOI: 10.1021/om100236v
Rhodium-Catalyzed Asymmetric Hydroformylation with Taddol-Based
IndolPhos Ligands
Jeroen Wassenaar, Bas de Bruin, and Joost N. H. Reek*
Supramolecular & Homogeneous Catalysis, van’t Hoff Institute for Molecular Sciences,
University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands
Received March 26, 2010
A small library of Taddol-based IndolPhos ligands 2a-g and their use in asymmetric hydro-
formylation (AHF) reactions are reported. Moderate to good enantioselectivities are obtained for
styrene, vinyl acetate, and allyl cyanide up to 72%, 74%, and 63% ee, respectively. High b/l ratios are
obtained, which results in a high net yield of the desired chiral aldehyde. An unprecedented
temperature-dependent reversal of enantioselectivity is found when using ligands 2d,e, which are
based on a xylyl-derived Taddol in the Rh-catalyzed AHF of styrene. Furthermore, these ligands give
the opposite enantiomer of the product for vinyl acetate and allyl cyanide when compared to the other
library members, all of which are based on (R,R)-tartaric acid. Ligands 2a and 2d display a similar
kinetic profile, which is best described by type I kinetics. Deuterioformylation experiments have
shown that insertion of the alkene into the Rh-H bond is irreversible under the conditions applied.
High-pressure NMR studies indicate that the hydridobiscarbonyl rhodium species, which is the
resting state of the catalyst, features equatorial-apical coordination of the ligands with a preference
for the phosphine on the apical position, isomer A. However, in the case of ligand 2d the isomer in
which the phosphoramidite occupies the apical position, B, is only marginally higher in free energy
than isomer A. It is proposed that formation of the product takes place via isomer B for ligands 2d,e
and via isomer A for all other ligands, explaining the observed reversal of enantioselectivity.
Introduction
center is introduced in this reaction, AHF represents a
promising tool for the production of optically active com-
pounds for pharmaceutical or agrochemical uses.3 AHF of
styrene derivatives gives rapid access to nonsteroidal anti-
inflammatory agents such as ibuprofen, ketoprofen, and
naproxen after oxidation of the aldehyde.3a When vinyl
acetate is used as a substrate, the resulting aldehydes are
conveniently transformed into R-amino acids, optically pure
isoxazolines, and imidazoles.4 However, applications to date
are rare due to the challenging nature of the reaction. As
opposed to other enantioselective reactions such as asym-
metric hydrogenation, where only activity and enantioselec-
tivity need to be controlled, AHF also demands a high degree
of regiocontrol to suppress the formation of linear alde-
hydes. Also, generally higher temperatures are required to
achieve sufficient reaction rates for AHF, but this often leads
to a decrease in regio- and enantioselectivity. Finally, the
aldehyde products are prone to racemization under certain
reaction conditions.
Hydroformylation of alkenes is arguably the most elegant
transition metal catalyzed transformation developed to date.
In a single reaction step, three substrates (alkene, CO, H2) are
combined to give highly valuable aldehyde products, which
may be further transformed by reduction (alcohols), oxida-
tion (acids/esters), or condensation (imines/amines). The
100% atom efficiency of the process has attracted much
industrial interest, and nowadays hydroformylation is the
largest industrial application of homogeneous catalysis.1 In
most of these processes linear aldehydes are desired, and
selective formation of these can be achieved by using wide-
bite-angle diphosphine ligands in the case of Rh-catalyzed
hydroformylation.2
Asymmetric hydroformylation (AHF) on the other hand
focuses on formation of the branched aldehydes. As a chiral
*Corresponding author. Fax: þ 31 20 5255604. Tel: þ 31 20 5256437.
E-mail: j.n.h.reek@uva.nl.
Over the past two decades a number of chiral phosphorus
ligands have been reported, which are able to address these
challenges to some extent. Most of these ligands rely on
relatively π-acidic phosphorus atoms such as phosphites,
phosphoramidites, phospholanes, and diazaphospholanes to
achieve good reaction rates as opposed to the electron-rich
phosphines most commonly used in asymmetric hydrogenation.
(1) For recent reviews on hydroformylation, see: (a) van Leeuwen,
P. W. N. M.; Claver, C. Rhodium Catalyzed Hydroformylation; Kluwer
Academic Publishers: Dordrecht, 2000; Vol. 22. (b) Breit, B.; Seiche, W.
Synthesis 2001, 1–36. (c) van Leeuwen, P. W. N. M., Homogeneous
Catalysis, Understanding the Art; Kluwer: Dordrecht, The Netherlands,
2004; pp 125-174.
(2) van Leeuwen, P. W. N. M.; Kamer, P. C. J.; Reek, J. N. H.;
Dierkes, P. Chem. Rev. 2000, 100, 2741–2769.
(3) For reviews on AHF, see: (a) Agbossou, F.; Carpentier, J. F.;
Mortreux, A. Chem. Rev. 1995, 95, 2485–2506. (b) Dieguez, M.; Pamies,
O.; Claver, C. Tetrahedron: Asymmetry 2004, 15, 2113–2122. (c) Klosin, J.;
Landis, C. R. Acc. Chem. Res. 2007, 40, 1251–1259.
(4) Thomas, P. J.; Axtell, A. T.; Klosin, J.; Peng, W.; Rand, C. L.;
Clark, T. P.; Landis, C. R.; Abboud, K. A. Org. Lett. 2007, 9, 2665–2668.
r
2010 American Chemical Society
Published on Web 05/27/2010
pubs.acs.org/Organometallics

2768 Organometallics, Vol. 29, No. 12, 2010
Wassenaar et al.
Figure 2. Structures of IndolPhos ligands 1a-f and Taddol-
based derivatives 2a,b.
were reported by the groups of van Leeuwen,11 Dieguez,12
Pizzano,13 Schmalz,14 and Reek.14,15
Our group developed hybrid phosphine-phosphoramidite
IndolPhos 1 and its Taddol-based derivatives 2 (Figure 2),16
which have been successfully applied in asymmetric hydro-
genation and allylic alkylation. The small bite angle of only
Figure 1. Structure of representative examples of successful
ligands for AHF.
Bidentate phosphite ligands such as (2R,4R)-chiraphite5 and
(S,S)-kelliphite6 give high regioselectivities but only moder-
ate ee (Figure 1). C2-Symmetric phospholane-type ligands,
e.g., (S,S)-Esphos developed by Wills et al., provide high
enantioselectivity for vinyl acetate but are unselective for
styrene.7 Landis and Klosin extended the success of phos-
pholane-type ligands with the introduction of diazapho-
spholanes, giving high ee’s for styrene, vinyl acetate, and
allyl cyanide.4,8
The most promising class of ligands for AHF are hybrid
ligands containing a phosphine and a phosphite or phosphor-
amidite. A major breakthrough was achieved by Takaya and
Nozaki with the development of Binaphos and its derivatives,
giving ee’supto94% fora range of substrates.9 Morerecently,
Zhang and co-workers reported even higher enantioselectivi-
ties with their phosphoramidite analogue of Binaphos, Yan-
Phos, giving up to 98% ee for styrene, vinyl acetate, and allyl
cyanide.10 However, both systems give rather low b/l ratios.
Other successful examples of phosphine-phosphite systems
16e
85° observed in the crystal structure of [Rh(1e)(cod)]BF4
prompted us to investigate the use of ligands 1 and 2 in Rh-
catalyzed AHF. By virtue of their small bite angle, these
ligands would efficiently enforce equatorial-apical coordina-
tion in the hydridobiscarbonyl rhodium intermediates, likely
leading to a high selectivity for the desired branched aldehyde.
Preliminary studies supported this, and high b/l ratios up to
17 along with good enantioselectivity up to 72% ee were
obtained.16a In this paper a full investigation is reported, focus-
ing on the Taddol-based derivatives 2. A reversal of enantio-
selectivity is observed by changing substituents on the Taddol
moiety. The origin of this effect has been investigated by kinetic
measurements, high-pressure NMR and IR spectroscopy, and
DFT calculations.
Results and Discussion
Ligand Synthesis. The Taddol-based IndolPhos deriva-
tives 2a-g are synthesized in a straightforward two-step
synthesis from the corresponding (R,R)-Taddol (Scheme 1).
Transformation of the diol into the phosphorochloridite is
achieved by treatment with PCl3 at low temperature under
basic conditions. The yields are quantitative, and the phos-
phorochloridites are used without further purification. Con-
densation with the corresponding indolylphosphine gives
IndolPhos derivatives 2a-g in good yield up to 68% after
flash SiO2 chromatography. The modularity of the synthetic
sequence allows for facile derivatization and fine-tuning of
the chiral pocket in the corresponding hydroformylation
catalysts. The steric and electronic properties have been
varied systematically on the phosphine part (2a-c) and the
Taddol moiety (2d-f), and the effect of substitution of the
indole ring on the 7-position has been examined using 2g. All
(5) (a) Babin, J. E.; Whiteker, G. T. Asymmetric Synthesis. World
Patent, WO 9303839, 1993. (b) Whiteker, G. T.; Briggs, J. E.; Babin, J. E.;
Barner, B. A. Asymmetric Catalysis Using Bisphosphite Ligands. In
Chemical Industries; Marcel Dekker, Inc.: New York, 2003; Vol. 89,
pp 359-367.
(6) (a) Cobley, C. J.; Gardner, K.; Klosin, J.; Praquin, C.; Hill, C.;
Whiteker, G. T.; Zanotti-Gerosa, A.; Petersen, J. L.; Abboud, K. A. J.
Org. Chem. 2004, 69, 4031–4040. (b) Cobley, C. J.; Klosin, J.; Qin, C.;
Whiteker, G. T. Org. Lett. 2004, 6, 3277–3280.
(7) (a) Breeden, S.; Wills, M. J. Org. Chem. 1999, 64, 9735–9738. (b)
Breeden, S.; Cole-Hamilton, D. J.; Foster, D. F.; Schwarz, G. J.; Wills, M.
Angew. Chem., Int. Ed. 2000, 39, 4106–4108.
(8) (a) Landis, C. R.; Jin, W. C.; Owen, J. S.; Clark, T. P. Angew.
Chem., Int. Ed. 2001, 40, 3432. (b) Clark, T. P.; Landis, C. R.; Freed, S. L.;
Klosin, J.; Abboud, K. A. J. Am. Chem. Soc. 2005, 127, 5040–5042.
(9) (a) Nozaki, K.; Sakai, N.; Nanno, T.; Higashijima, T.; Mano, S.;
Horiuchi, T.; Takaya, H. J. Am. Chem. Soc. 1997, 119, 4413–4423. (b)
Nozaki, K.; Takaya, H.; Hiyama, T. Top. Catal. 1997, 4, 175–185.
(10) (a) Yan, Y.; Zhang, X. J. Am. Chem. Soc. 2006, 128, 7198–7202.
(b) Zhang, X.; Cao, B.; Yan, Y.; Yu, S.; Ji, B.; Zhang, X. Chem.;Eur. J.
2010, 16, 871–877.
(14) Robert, T.; Abiri, Z.; Wassenaar, J.; Sandee, A. J.; Romanski, S.;
Neudorfl, J.-M.; Schmalz, H.-G.; Reek, J. N. H. Organometallics 2010,
29, 478–483.
(11) (a) Deerenberg, S.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.
Organometallics 2000, 19, 2065–2072. (b) Deerenberg, S.; Pamies, O.;
Dieguez, M.; Claver, C.; Kamer, P. C. J.; van Leeuwen, P. W. N. M. J. Org.
Chem. 2001, 66, 7626–7631.
(12) Dieguez, M.; Pamies, O.; Ruiz, A.; Castillon, S.; Claver, C.
Chem.;Eur. J. 2001, 7, 3086–3094.
(15) Chikkali, S. H.; Bellini, R.; Berthon-Gelloz, G.; van der Vlugt,
J. I.; de Bruin, B.; Reek, J. N. H. Chem. Commun. 2010, 46, 1244–1246.
(16) (a) Wassenaar, J.; Reek, J. N. H. Dalton Trans. 2007, 3750–3753.
(b) Wassenaar, J.; Kuil, M.; Reek, J. N. H. Adv. Synth. Catal. 2008, 350,
1610–1614. (c) Wassenaar, J.; Reek, J. N. H. J. Org. Chem. 2009, 74, 8403–
8406. (d) Wassenaar, J.; van Zutphen, S.; Mora, G.; Le Floch, P.; Siegler,
M. A.; Spek, A. L.; Reek, J. N. H. Organometallics 2009, 28, 2724–2734.
(e) Wassenaar et al. Chem. Eur. J. 2010, DOI: 10.1002/chem.200903476.
(13) Rubio, M.; Suarez, A.; Alvarez, E.; Bianchini, C.; Oberhauser,
W.; Peruzzini, M.; Pizzano, A. Organometallics 2007, 26, 6428–6436.

Article
Organometallics, Vol. 29, No. 12, 2010 2769
Scheme 1. Synthesis of Taddol-Based IndolPhos Ligands 2a-g
Table 1. Screening of Ligands in AHF of Styrenea
ligands are easy to handle as white powders and are stable as
a solid toward air and moisture.
Similar to the bisnaphthol-based IndolPhos ligands 1, the
Taddol-based derivatives display large coupling constants
between the two inequivalent phosphorus atoms in the 31P
NMR spectra (JPP ≈ 200 Hz). In the case of isopropyl phos-
phines (2b and 2e), two conformations of the ligand are
observed at low temperatures of which only one displays a
P-P coupling. We believe that this is due to rotation of the
phosphine-carbon bond, leading to two conformers where
the phosphorus lone pairs point toward each other or away
from one another. In the former case a large coupling con-
stant is expected but not in the latter.
Ligand Screening in AHF. Preliminary experiments in the
AHF of styrene using IndolPhos ligands 1 showed that only the
isopropyl-substituted ligands 1b,f were giving appreciable levels
of activity and regio- and enantioselectivity.16a In the case of
phenyl-substituted ligands no conversion was obtained, which is
likely explained by the formation of bis-ligated species, as has
been observed for TangPhos systems.16a,17 The more bulky
Taddol-based ligands should not suffer from this problem,
and indeed active catalysts are generated with all ligands.
Styrene was used as benchmark substrate, and ligand screening
was carried out under standard conditions (20 bar of CO/H2,
PhMe, L/Rh = 2.0) at 40 and 70 °C (Table 1). Even though
higher rates can be expected at lower syngas pressures, we chose
20 bar, as lower pressures may have a detrimental effect on the
enantioselectivity and conversion due to slow formation of the
active species (vide infra).
Taddol-based ligand 2a provides the hydroformylation
product of styrene in 71% ee, which is almost identical to
bisnaphthol-based ligand 1b (entries 3 and 5). However, the
regioselectivity is much higher (b/l ratio of 28) and the
conversion somewhat lower. Increasing the steric bulk on
the phosphine moiety leads to a decrease in activity, regio-
selectivity, and enantioselectivity (entries 7 and 9). The use of
more bulky xylyl paddlewheels on the Taddol fragment leads
to a decrease in activity and enantioselectivity, but raises the
b/l ratio up to 37 (entries 11 and 13). At 40 °C, only traces of
hydroformylation products were obtained employing the
perfluorated ligand 2f (entry 15). We found previously that
substituents on the ligand backbone of Taddol-containing
phosphine-phosphite ligands dramatically enhanced the
enantioselectivity in the AHF of styrene.14 However, ligand
entry
ligand
T (°C)
% convb
b/lc
% eed
1
2
3
4
5
6
7
8
1a
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
32
99
96
100
57
100
3
69
8
98
48
100
8
14
6
10
6
64 (R)
46 (R)
72 (R)
53 (R)
71 (R)
46 (R)
55 (R)
44 (R)
48 (R)
42 (R)
33 (R)
16 (S)
37 (R)
10 (S)
1b
2a
2b
2c
2d
2e
2f
28
14
14
10
2
21
37
20
17
5
9
10
11
12
13
14
15
16
17
18
100
<1
98
10
100
21
5
6
11 (S)
58 (R)
3 (R)
2g
a [Rh(acac)(CO)2] = 1.0 mmol/L in toluene, [ligand] = 2.0 mmol/L,
styrene/rhodium = 1000, pressure = 20 bar (CO/H2 = 1:1). b Percen-
tage conversion determined by GC; the reaction was stopped after 20 h.
c Ratio of branched to linear product. d Enantiomeric excess determined
by chiral GC (Supelco BETA DEX 225).
2g, containing a methyl group in the 7-position of the indole
backbone, gives a lower activity, b/l ratio, and ee (entry 17).
When the temperature is raised to 70 °C, almost all ligands
provide catalysts that give full conversion, but both enan-
tioselectivity and regioselectivity are significantly lower.
Remarkably, raising the temperature from 40 to 70 °C, when
applying ligands 2d and 2e, reverses the enantioselection
from R (37% ee) to S (up to 16% ee, entries 11-14). This
result is highly unexpected and has up to now been observed
only in Pt/Sn-catalyzed AHF reactions.18 It may be caused
by the presence of multiple active isomers of the catalyst,
with a temperature dependent relative distribution and/or a
different response of these isomers in rate to temperature
changes. The phenomenon will be discussed in more detail
below.
(18) (a) van Duren, R.; Cornelissen, L. L. J. M.; van der Vlugt, J. I.;
€
Huijbers, J. P. J.; Mills, A. M.; Spek, A. L.; Muller, C.; Vogt, D. Helv.
ꢀ
Organometallics 1993, 12, 848–852.
(17) Axtell, A. T.; Klosin, J.; Abboud, K. A. Organometallics 2006,
25, 5003–5009.
Chim. Acta 2006, 89, 1547–1558. (b) Toth, I.; Guo, I.; Hanson, B. E.

2770 Organometallics, Vol. 29, No. 12, 2010
Wassenaar et al.
Table 2. Screening of Ligands in AHF of Vinyl Acetatea
Table 3. Screening of Ligands in AHF of Allyl Cyanidea
entry
ligand
T (°C)
% convb
b/lc
1.2
16.8
4.2
777
21.7
40
% eed
entry
ligand
T (°C)
% convb
b/lc
% eed
1
2
3
4
5
6
7
8
1a
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
2
100
5
24 (S)
19 (S)
31 (S)
42 (S)
72 (S)
33 (S)
1
2
3
4
5
6
7
8
1a
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
40
70
10
3
50
51 (S)
45 (S)
63 (S)
55 (S)
59 (S)
50 (S)
40 (S)
36 (S)
45 (S)
43 (S)
35 (R)
29 (R)
47 (R)
39 (R)
24 (S)
17 (S)
37 (S)
3 (S)
16.5
8.4
6.3
4.4
3.2
4.8
2.1
1.2
1.0
4.1
3.3
4.1
2.3
8.4
2.3
3.5
2.4
1b
2a
2b
2c
2d
2e
2f
1b
2a
2b
2c
2d
2e
2f
78
99
71
99
6
96
12
98
39
99
8
98
4
98
14
99
97
19
98
<1
49
<1
87
13
99
2
199
36 (S)
9
9
10
11
12
13
14
15
16
17
18
25.7
8
97
0.3
59
6 (R)
10
11
12
13
14
15
16
17
18
59 (R)
35 (R)
39 (R)
29 (R)
49
<1
99
2
1.5
0.3
15.1
10 (S)
51 (S)
<2
2g
2g
100
a [Rh(acac)(CO)2]=1.0 mmol/L in toluene, [ligand]=2.0 mmol/L, vinyl
acetate/rhodium = 1000, pressure = 20 bar (CO/H2 = 1:1). b Percentage
conversion determined by GC; the reaction was stopped after 20 h. c Ratio of
branched to linear product. d Enantiomeric excess determined by chiral GC
(Supelco BETA DEX 225).
a [Rh(acac)(CO)2]=1.0 mmol/L in toluene, [ligand]=2.0 mmol/L, allyl
cyanide/rhodium = 1000, pressure =20 bar (CO/H2 =1:1). b Percentage
conversion determined by GC; the reaction was stopped after 20 h. c Ratio of
branched to linear product. d Enantiomeric excess determined by chiral GC
(Astec Chiraldex A-TA).
63% ee for styrene, vinyl acetate, and allyl cyanide, respectively.
Using the full library of IndolPhos ligands 1 and 2, the substrate
specificity of the generated catalysts is low, making them
attractive for general use. The enantioselectivities are not among
the highest reported but certainly useful as the ee of commer-
cially relevant products may be increased by crystallization.
Especially the high b/l ratios result in a high net yield of the
desired optically active branched aldehyde. The dependence of
the absolute configuration of the products on the temperature
and substituents on the ligand is unexpected and up to now
unexplored in the literature. Understanding these effects may
lead to a better mechanistic picture and eventually a route to
rational optimization of the catalysts. Therefore kinetic studies,
high-pressure NMR experiments, and DFT calculations were
performed to uncover the origin of the enantioreversal.
Kinetics. In order to determine whether the catalysts
generated from ligands 2a and 2d follow the same catalytic
cycle, we monitored the progress of the AHF of vinyl acetate
in time using in situ gas uptake measurements. The rate-
determing step in hydroformylation reaction can be early or
late in the catalytic cycle, leading to type I or type II kinetics,
respectively (eqs 1 and 2).
To study the scope, we applied IndolPhos ligands 1 and 2 in
the AHF of vinyl acetate (Table 2). Bisnaphthol-based
ligands 1a,b yield unreactive and unselective catalysts for this
substrate, as conversions of only up to 5% at low b/l ratios
and ee are achieved (entries 1 and 3). Conversely, Taddol-
based ligand 2a leads to moderate rates and a good ee of 72%
and b/l ratio of 22, similar to the performance in the case of
styrene (entry 5). Surprisingly, the xylyl-functionalized ligand
2d gives the opposite enantiomer in 59% ee (entry 11, Δee=
141%). Both ligands contain the same absolute configuration
of the Taddol moiety (R,R) and differ only in substituents on
the paddlewheels. Such a dramatic reversal of absolute con-
figuration induced by a slight modification of the substituents
on the ligand in AHF reactions has been observed only for
Josiphos ligands (Δee e82%).17 The other ligands display
very low activities at 40 °C. At 70 °C, most ligands give
quantitative conversion, except for the more bulky isopropyl-
and o-tolyl-functionalized ligands 2b, 2c, and 2e. The b/l ratio
increases at higher temperatures, which in some cases leads to
almost exclusive formation of the branched product with a b/l
ratio up to 777 (entries 4 and 8). The enantioselectivity is
generally lower at higher temperature, but remarkably, the
application of ligand 1b at higher temperature results in an
increase of ee (from 31% to 42%, entry 4).
AHF of allyl cyanide is a potentially valuable process, as it
furnishes β-amino acid derivatives after hydrogenation and
oxidation. Application of IndolPhos ligands 1 and 2 in this
reaction lead to b/l ratios up to 50% and 63% ee at high
conversion (Table 3). In general b/l ratios for this substrate are
low, and the regioselectivity provided by ligand 1a is exceptional
and among the highest reported to date. The absolute config-
uration also reverses for this substrate when using the more
sterically demanding xylyl-functionalized ligands 2d and 2e,
similar to the case of vinyl acetate (entries 11-14).
A½alkeneꢀ½Rhꢀ
rateðtype IÞ ¼
ð1Þ
B þ ½Lꢀ
C½H2ꢀ½Rhꢀ
D þ ½COꢀ
rateðtype IIÞ ¼
ð2Þ
(A-D are constants, [L] is proportional to the [phosphine
ligand] and [CO]).
In type I kinetics, positive orders are found in [alkene] and
[rhodium], a zero order in pH2, and a negative order in the
[ligand] and/or pCO. Most catalyst systems follow these
kinetics, which is consistent with a rate-determining alkene
Summarizing, the ligand screening experiments in the AHF of
three substrates gave enantioselectivities up to 72%, 72%, and

Article
Organometallics, Vol. 29, No. 12, 2010 2771
Table 4. Influence of Syngas Pressure on the Asymmetric
Hydroformylation of Vinyl Acetate with Ligands 2a and 2da
To investigate whether insertion of the alkene is reversible,
we carried out deuterioformylation experiments with vinyl
acetate using a CO/D2 (1:1) gas mixture. Under standard
conditions similar conversions and higher b/l ratios are
observed (Scheme 2). The pathway leading to the linear
aldehyde is less favorable when deuterium is used, as was
also observed for Binaphos by Nozaki and Takaya.19 The ee
is not significantly affected by the use of D2 instead of H2.
Analysis of the products by 1H NMR shows that deuterium
is incorporated only at the expected positions, i.e., on the
aldehyde and methyl groups. Since no scrambling of the
deuterium atoms occurs, the insertion of the alkene in the
Rh-H bond is irreversible under the applied conditions.19
High-Pressure NMR and IR Spectroscopy. The hydrido-
biscarbonyl species of all Taddol-based IndolPhos ligands 2
were studied by high-pressure 1H and 31P NMR spectrosco-
py. As a reference, bisnaphthol-based ligand 1a was also used
in this study. Three different isomers can be expected for
unsymmetrical hybrid ligands, the ee (bis-equatorial) isomer
and two ea (equatorial-apical) isomers, A and B, in which
either the phosphine or the phosphoramidite occupies the
apical position (Figure 4). The complexes were generated in a
high-pressure NMR tube from the corresponding acac com-
plex that was exposed to 5 bar of CO/H2 at 50 °C for 16 h.
The chemical shifts of the hydride and phosphorus signals as
well as the coupling constants with the hydride are depicted
in Table 5. Ligand 2c gave rise to the formation of a new
species under syngas atmosphere observed by 31P NMR;
however, no hydride signal could be detected. In the case of
ligand 2g only decomposition was observed after prolonged
heating under syngas atmosphere.
entry ligand PCO/H2 (bar) % convb TOF (h-1 c
)
b/ld % eee
1
2
3
4
5
6
7
8
2a
10
20
30
40
10
20
30
40
51
23
20
16
37
23
15
11
7.6
3.5
3.1
2.5
6.5
3.5
2.7
2.2
58.7 73 (S)
19.0 73 (S)
14.3 74 (S)
9.9 73 (S)
21.4 58 (R)
12.0 59 (R)
9.6 60 (R)
5.4 59 (R)
2d
a [Rh(acac)(CO)2] = 1.0 mmol/L in toluene, [ligand] = 2.0 mmol/L, vinyl
acetate/rhodium = 200, temperature = 40 °C. b Percentage conversion
determined by GC; the reaction was stopped after 17 h. c Turnover frequency
determined in the linear part of the reaction profile (10-20% conv). d Ratio
of branched to linear product. e Enantiomeric excess determined by chiral
GC (Supelco BETA DEX 225).
coordination or migratory insertion of the alkene in the
Rh-H bond. Type II kinetics has been reported for bulky
phosphite systems, giving fast catalysts in which oxidative
addition of H2 to the acylrhodium species is rate determin-
ing, hence a positive order in pH2. In type I kinetics the
enantioselectivity is determined in the migratory insertion
under conditions where this step is irreversible.1a,19
Table 4 summarizes the performance of ligands 2a and 2d
in the hydroformylation of vinyl acetate. The enantioselec-
tivity is independent of the pressure, whereas the b/l ratio is
lower at higher syngas pressures. The latter is unexpected, as
usually higher CO pressures result in higher b/l ratios.1a Also
the conversion is lower at higher pressures, thus indicating a
negative order in pCO.
The metal complexes of all ligands (except for 2c and 2g)
give rise to either a doublet-of-doublets-of-doublets or a
doublet-of-triplets (in case one of the hydride-phosphorus
1
couplings equals the Rh-H coupling constant) in the H
NMR spectrum at high field (-8-10 ppm). No other
hydride signals were detected under the indicated conditions.
At least one of the hydride-phosphorus couplings is very
large, which is indicative for a trans disposition of hydride
and the phosphorus atom. This observation already excludes
formation of ee complexes, which is in line with what we
expected on the basis of the small bite-angle of the ligand.
The Rh-H coupling constant is around 8-10 Hz, typical for
trigonal-bipyramidal [Rh(H)(CO)2(L)2] complexes.
The lower rates at higher pressures are also reflected in the
reaction profiles derived from gas uptake measurements
(Figure 3). Turnover frequencies are obtained from these
profiles, which are given in Table 4. A significant decrease of
the rate is observed with increasing pressure, again in line
with the negative order in pCO. At 10 bar the curves show an
induction period up to 4 h, most likely resulting from slow
formation of the active hydridobiscarbonyl species from the
acac complex. At higher pressures, the active species is
formed faster and no induction period is observed. After
the induction period, a constant slope is observed for both
ligands, indicating steady-state kinetics up to at least 50%
conversion. This can be explained by reversible alkene
coordination prior to an irreversible migratory insertion,
which results in a pseudo-zero-order in [alkene] at high
concentrations of the substrate. Ligand 2a gives a slightly
higher rate compared to ligand 2d, an effect that is more
pronounced at lower pressures.
The kinetics observed for both systems above are in best
agreement with type I kinetics, as we found a strong decrease
in rate with increasing syngas pressure. The rate-determining
step is proposed to be the migratory insertion because
steady-state kinetics are observed in the beginning of the
reaction, indicating reversible alkene coordination. The
absolute configuration is determined in the insertion step,
if this step is irreversible.
Bisnaphthol-based ligand 1a exhibits large values for the
2JHP(phosphine) and 2JHP(phosphoramidite) couplings, in which the
latter is the largest one (Table 5, entries 1, 2). This indicates a
rapid equilibrium between isomers A and B, in which B is the
2
major species.20 The temperature dependence of the JHP
couplings confirms this proposition. For the Taddol-based
ligands, an opposite trend is observed with large values for
2JHP(phosphine), indicating a preference for isomer A (Table 5,
entries 3-14). The temperature-dependent coupling con-
stants allow calculation of the ratio between A and B, which
is >99:1 at 45 °C.20 The more bulky xylyl-functionalized
Taddol in ligands 2d and 2e gives rise to slightly larger values
of 2JHP(phosphoramidite) compared to the ligands containing the
parent Taddol. This indicates that the equilibrium between A
and B is slightly shifted toward isomer B for the more bulky
ligand 2d (A/B = 95:5 at 45 °C), which is also reflected in the
(20) Casey, C. P.; Paulsen, E. L.; Beuttenmueller, E. W.; Proft, B. R.;
Petrovich, L. M.; Matter, B. A.; Powell, D. R. J. Am. Chem. Soc. 1997,
119, 11817–11825.
(19) Horiuchi, T.; Shirakawa, E.; Nozaki, K.; Takaya, H. Organo-
metallics 1997, 16, 2981–2986.

2772 Organometallics, Vol. 29, No. 12, 2010
Wassenaar et al.
Figure 3. Reaction profiles of the asymmetric hydroformylation of vinyl acetate using ligands 2a and 2d at 10, 20, 30, and 40 bar of
CO/H2 and 40 °C. Conversions are obtained from gas uptake curves using the ideal gas law and are uncorrected. Dissolution of the gas
into the solution is not entirely finished at the point of reaction start, leading to slightly higher conversions than obtained by GC.
Scheme 2. Deuteroformylation of Vinyl Acetate with Ligands
2a and 2d
Figure 4. Structures of bis-equatorial isomer ee and equatorial-
apical isomers A and B of the hydridobiscarbonyl rhodium species.
smaller energy difference between these species found by
DFT calculations (vide infra).
For hybrid phosphine-phosphoramidite ligands, no high-
pressure NMR studies of the hydridobiscarbonyl rhodium
species have been reported so far. For phosphine-phosphites
on the other hand, examples have been reported with a pre-
ference both for isomer A (van Leeuwen,11a Claver,21 Pizzano,13
Schmalz and Reek14) and for isomer B (Binaphos9a). The
presence of both isomers A and B in the case of IndolPhos
ligands 2a and 2d raises the question, which will be the active
species leading to the product? In analogy to the mechanism of
asymmetric hydrogenation, the minor isomer Bmay be far more
active, resulting from a lower transition-state barrier for alkene
insertion than A and thus leading to the product.
High-pressure in situ IR spectroscopy enables the study of the
catalyst resting state during the reaction, giving valuable me-
chanistic information.22 We studied the formation of the
hydridobiscarbonyl rhodium species for ligands 2a and 2d,
which were formed within 20 h at 40 °C and 20 bar of CO/H2 in
cyclohexane (Figure 5). As expected, only one set of signals
Table 5. NMR Data for the Hydride Resonance of
RhH(IndolPhos)(CO)2 Complexesa
entry ligand T (°C) δHRh 1JRhH δPC 2JHPC δPN 2JHPN
1
2
3
4
5
9
10
11
12
13
14
1a
2a
25
45
25
45
25
25
45
25
45
25
45
-8.40
-8.48
-8.78
-8.82
-8.35
-8.62
-8.69
9.9
9.7
9.5
9.4
9.1
9.6
9.6
68.2
68.5
66.7 176.7 95.7
63.8 176.9 101.9
29.7 113.1 144.0
29.8 112.7 143.8
3.6
4.8
2b
2d
69.3 109.1 142.5 ndb
29.5 110.5 142.2
29.5 109.2 142.2
69.1 108.4 140.4
69.1 107.6 140.4
9.6
12.3
6.2
8.7
2.3
2e
2f
-8.26 10.0
-8.34
-9.64
-9.67
8.7
9.7 ndc
8.6 ndc
111.0 ndc
112.3 ndc
ndb
a Hydridobiscarbonyl complexes were generated from [Rh(acac)-
(CO)2] and the corresponding ligand by heating for 16 h at 50 °C under
5 bar of CO/H2 in C6D6. b Not detected due to zero or near zero value of
coupling constant. c Not detected due to low concentration.
(21) Pamies, O.; Net, G.; Ruiz, A.; Claver, C. Tetrahedron: Asym-
metry 2002, 12, 3441–3445.
(22) Kamer, P. C. J.; van Rooy, A.; Schoemaker, G. C.; van Leeuwen,
(symmetric and antisymmetric stretch) for the carbonyls of the
ea complex is observed in both cases. The calculated IR absorp-
tions (ri-DFT, BP86, SV(P)) of isomer A for ligand 2a (2016.9
P. W. N. M. Coord. Chem. Rev. 2004, 248, 2409–2424.

Article
Organometallics, Vol. 29, No. 12, 2010 2773
Figure 5. High-pressure IR spectra of RhH(IndolPhos)(CO)2 complexes of ligands 2a (left) and 2d (right) after incubation (black) and
in the presence of vinyl acetate (red).
Table 6. Calculated Energies of Hydridobiscarbonyl Rhodium
Isomers A and B for Ligands 2a and 2d at the ri-DFT BP86 Level
Using the SV(P) Basis Set on All Atomsa
b
entry ligand isomer T (°C) ΔEZPE
ΔGc
ΔHd
ΔSe
1
2
3
4
5
6
7
8
2a
A
B
A
B
A
B
A
B
25
25
40
40
25
25
40
40
0.0
3.1
0.0
3.1
0.0
2.5
0.0
2.5
0.0
3.4
0.0
3.3
0.0
0.1
0.0
-0.1
0.0
3.3
0.0
3.3
0.0
2.4
0.0
2.5
0.0
-0.5
0.0
-0.2
0.0
7.8
2d
0.0
8.5
a All energies are relative to isomer A for each ligand at each
temperature. b Zero point energy corrected SCF energy in kcal/mol.
c Standard free energy difference in the gas phase (1 bar), in kcal/mol.
d Standard enthalpy difference in the gas phase in kcal/mol. e Standard
entropy difference in the gas phase in cal/mol/K.
dormant species that is in equilibrium with the active species. In
line with this, the signal disappears slowly as the reaction
proceeds. The nature of this dormant species is currently not
clear.
DFT Calculations. In order to gain a more detailed under-
standing of the relative energies of isomers A and B for
ligands 2a and 2d, and to assign the IR spectra more
accurately, we carried out DFT calculations on these com-
plexes. The calculated structures for the RhH(IndolPhos)-
(CO)2 species of ligands 2a and 2d are displayed in Figure 6a,
b, respectively. The energies are summarized in Table 6.
The geometries obtained after optimization for isomers Aand
B are very similar for both ligands. The increased steric bulk of
the xylyl-substituted ligand 2d does not lead to a large con-
formational change. Conversely, the steric bulk of 2d has a
major impact on the stability of the two isomers, as the energy
difference between A and B is much smaller for ligand 2d
compared to 2a. For 2a, isomer A is 3.4 kcal/mol more stable,
which is also reflected in the small fraction of B(<1%) found by
HP-NMR. For 2d, isomer B is almost equal in free energy to A.
This is in good agreement with the HP-NMR studies, where
Figure 6. DFT-calculated structures of isomers A and B with
(a) ligand 2a and (b) ligand 2d.
and 1990.2 cm-1) and 2d (2012.0 and 1984.6 cm-1) correspond
well with the experimentally found absorptions (2a: 2022.5 and
1984.3 cm-1; 2b: 2017.7 and 1979.2 cm-1). Visualization of the
computed CO vibrations of isomer A for both ligands reveals
that they are strongly coupled with the Rh-H stretching
modes. The observation of only one set of signals indicates
that the carbonyl ligands of isomer A dominate the spectrum.
Isomer B seems to be visible in small amounts as red-shifted
shoulders partly overlapping the main signals of isomer A. The
shoulders are more pronounced for 2d, in agreement with the
NMR data. Upon addition of vinyl acetate, a second species
forms in both experiments at slightly higher wavenumbers
((2054 cm-1), exhibiting only a single peak. This signal is
unlikely to stem from a monocarbonyl alkene adduct, [RhH-
(2a)(CO)(vinyl acetate)]. Calculation of those complexes gave
IR absorptions at much lower wavenumbers, as is to be expected.
Instead, it seems that the substrate induces the formation of a
2
larger values for JHP(phosphoramidite) were found in the case of
ligand 2d, and about 5% of isomer B is present in solution at
45 °C. The slightly larger energy difference found spectroscopi-
cally may be attributed to solvent effects. Moreover, a positive
entropy contribution is observed for conversion of A (2d) to B
(2d) versus a negative contribution in the case of conversion of A

2774 Organometallics, Vol. 29, No. 12, 2010
Wassenaar et al.
Scheme 3. Mechanism of Enantioselection of Ligands 2a (left) and 2d (right) Leading to the Opposite Enantiomers of the Product
(2a) to B (2a). This may explain the temperature-dependent
reversal of enantioselectivity when styrene is used as a substrate
(Table 1).
tivity using ligands 2d-e, which are based on a xylyl-derived
Taddol in the Rh-catalyzed AHF of styrene. Furthermore, these
ligands give the opposite enantiomer of the product for vinyl
acetate and allyl cyanide when compared to the other library
members, all of which are based on (R,R)-tartaric acid.
Proposed Mechanism of Enantiodiscrimination. On the
basis of the results outlined above, it can be envisioned that
the reversal of enantioselectivity observed in the asymmetric
hydroformylation using ligands 2a and 2d is a result of the
difference in relative stability of ea isomers A and B for ligands
2a and 2d. The migratory insertion of the alkene into the Rh-H
bond, which is proposed to be the rate- and selectivity-determin-
ing step (vide supra), is accompanied by a rotation of the double
bond from in-plane coordination to a perpendicular mode in the
transition state. For ligand 2a the reaction may proceed via
hydridobiscarbonyl rhodium species A, which, after loss of one
molecule of CO, selectively coordinates the alkene, which inserts
into the Rh-H bond in TS A (Scheme 3). The configuration of
the phenyl rings on the phosphine, controlled by the chiral
information in the Taddol moiety, blocks the double-bond
rotation selectively, similarly to that proposed by Pizzano
et al.13 In the case of ligand 2d the main reaction flux may be
carried by isomer B and the selectivity is determined in TS B.
One of the xylyl paddlewheels of the Taddol moiety selectively
blocks one coordination site. The other site can selectively
discriminate between the two prochiral faces of the olefin.
However, when the substituent on the olefin becomes larger,
as is the case in styrene, the substrate does not fit very well, and
reaction via TS A becomes competitive. Only by shifting the
equilibrium to isomer B using higher temperatures (ΔS is
positive) can TS B dominate the outcome of the reaction.
The proposed mechanism assumes similar barriers for
alkene insertion for both isomers A and B. However, it is
known that in many reactions the minor species may lead to
the product, like in asymmetric hydrogenation. Therefore,
calculation of these transition states is necessary to give a
definitive answer. Nonetheless, the proposed mechanism is
in accordance with all data obtained by experimental and
theoretical techniques.
We investigated the origin of this enantioreversal by kinetic
studies, high-pressure NMR and IR, and DFT calculations. It
was found that ligands 2a and 2d reveal a similar kinetic profile,
which is best described by type I kinetics. Deuterioformylation
experiments have shown that insertion of the alkene into the
Rh-H bond is irreversible under the conditions applied. The
hydridobiscarbonyl rhodium species, which is the resting state of
the catalyst, features equatorial-apical coordination of the
ligands with a preference for the phosphine on the apical
position, isomer A. However, DFT calculations and high-pres-
sure NMR studies indicate that in the case of ligand 2d the
isomer in which the phosphoramidite occupies the apical posi-
tion, B, is only marginally higher in free energy than isomer A.
We therefore propose that formation of the product takes place
via isomer B for ligands 2d-e and via isomer A for all other
ligands, explaining the observed reversal of enantioselectivity.
In conclusion, we have demonstrated the efficient applic-
ability of catalysts derived from Taddol-based IndolPhos
ligands in AHF reactions. Moreover, the detailed investiga-
tion triggered by an unprecedented reversal of enantioselec-
tivity has provided valuable insights into the mechanism of
enantiodiscrimination in AHF reactions that may also be
applicable for other catalytic systems.
Experimental Section
General Procedures. All reactions were carried out under an
atmosphere of argon using standard Schlenk techniques. THF,
pentane, hexane, and diethyl ether were distilled from sodium
benzophenone ketyl; CH2Cl2, 2-propanol, and methanol were
distilled from CaH2; and toluene was distilled from sodium under
nitrogen. With the exception of the compounds given below, all
reagents were purchased from commercial suppliers and used with-
out further purification. The following compounds were synthesized
according to published procedures: substituted and unsubstituted
Taddols,23 diphenyl(3-methyl-2-indolyl)phosphine,16a diisopropyl-
(3-methyl-2-indolyl)phosphine,16a IndolPhos ligands 1a-f,16a,b
Taddol-based IndolPhos ligands 2a,b,16c 3,7-dimethylindole.24
Melting points were recorded on a Gallenkamp melting point
Conclusions
Summarizing, we have reported the synthesis of a small library
of Taddol-based IndolPhos ligands and their use in AHF
reactions. Moderate to good enantioselectivities are obtained
for styrene, vinyl acetate, and allyl cyanide up to 72%, 74%, and
63% ee, respectively. High b/l ratios are obtained, which results
in a high net yield of the desired chiral aldehyde. We found an
unprecedented temperature-dependent reversal of enantioselec-
(23) Seebach, D.; Beck, A. K.; Heckel, A. Angew. Chem., Int. Ed.
2001, 40, 92–138.
(24) Dobbs, A. P.; Voyle, M.; Whittall, N. Synlett 1999, 1594–1596.

Article
Organometallics, Vol. 29, No. 12, 2010 2775
apparatus and are uncorrected. NMR spectra (1H, 31P, and 13C)
were measured on a Varian INOVA 500 MHz. Optical rotations
were determined on a Perkin-Elmer 241 polarimeter. High-
resolution mass spectra were recorded on a JEOL JMS SX/
SX102A four-sector mass spectrometer; for FAB-MS 3-nitro-
benzyl alcohol was used as matrix. Chiral GC separations were
conducted on an Interscience Focus GC (FID detector) with a
Supelco BETA DEX 225 column (0.25 mm ꢁ 30 m) or an Astec
Chiraldex A-TA column (0.25 mm ꢁ 30 m).
J = 215.7 Hz, 0.6P), -39.15 (d, J = 216.3 Hz, 0.6P), -42.14 (d,
J = 218.2 Hz, 0.4P) ppm. HRMS (FAB): calcd for [M þ H]þ
C54H50NO4P2, 838.3215; found, 838.3207.
Synthesis of Taddol-Based IndolPhos Ligand 2d. To a solution
of diphenyl(3-methyl-2-indolyl)phosphine (0.209 g, 0.66 mmol) in
THF (5 mL) was added dropwise 0.28 mL of n-BuLi (2.5 M in
hexanes) at -78 °C. The resulting yellow solution was stirred for
0.5 h at -78 °C, after which a solution of the phosphorochloridite
derived from (4R,5R)-2,2-dimethyl-R,R,R0,R0-tetrakis(3,5-dimethyl-
phenyl)-1,3-dioxolane-4,5-dimethanol (0.427 g, 0.66 mmol) in THF
(5 mL) was added. The reaction mixture was stirred overnight,
allowing it to warm to room temperature. The resulting pale yellow
solution was evaporated to dryness, and the resulting residue was
purified by flash SiO2 column chromatography (3% EtOAc in
hexane) to give a white solid. Yield: 0.372 g (61%). R2D2 = þ97.6 (c
1.0, CHCl3). Mp = 163.4 °C. 1H NMR (500 MHz; CDCl3; 298 K):
δ 8.84 (d, J = 8.3 Hz, 1H), 7.64 (d, J = 7.9 Hz, 1H), 7.52 (t, J = 7.7
Hz, 1H), 7.44 (s, 2H), 7.39-7.26 (m, 14H), 7.16 (s, 4H), 6.82 (d, J =
7.3Hz, 2H), 6.76(s, 1H), 5.41(dd, J= 8.2, 2.9 Hz, 1H), 5.00 (d, J=
8.4 Hz, 1H), 2.27 (s, 6H), 2.22 (s, 6H), 2.13 (s, 6H), 2.07 (s, 6H), 1.75
(s,3H),1.66(s,3H),0.39(s,3H) ppm.13C NMR (126 MHz; CDCl3;
298 K): δ 146.3, 145.0, 141.2, 140.9, 139.0, 137.3 (d, JCP = 8.0 Hz),
General Procedure for the Synthesis of Taddol Phosphoro-
chloridites. The desired Taddol was azeotropically dried prior to
use by co-evaporation with PhMe. To a solution of Et3N (4.00
mmol) in THF (10 mL) was added subsequently PCl3 (2.08
mmol) and a solution of the corresponding Taddol (1.98 mmol)
in THF (10 mL) at -78 °C. The mixture was stirred for 30 min at
-78 °C and then allowed to warm to rt. The resulting white
suspension was filtered over Celite, and the filtrate was concen-
trated in vacuo to obtain a white solid in quantitative yield,
which was used without further purification.
Synthesis of Diphenyl(3,7-dimethyl-2-indolyl)phosphine. To a
solution of 3,7-dimethylindole (0.93 g, 6.4 mmol) in THF (20 mL)
was added dropwise 2.7 mL of n-BuLi (2.5 M in hexanes) at -78 °C.
The resulting deep red solution was stirred at -78 °C for 20 min.
Carbon dioxide was bubbled through the mixture for 10 min, which
was allowed to warm to room temperature, and the solvent was
removed in vacuo. The resulting residue was dissolved in THF (20
mL) to give a clear solution, which was cooled to -78 °C. To this
solution was added 4.2 mL of t-BuLi (1.6 M in pentanes), and the
resulting orange solution was stirred at -78 °C for 1 h. Chlorodi-
phenylphosphine (1.15 mL, 6.4 mmol) was added, and the reaction
mixture was stirred for 16 h, allowing it to warm to room
temperature. The resulting yellow solution was washed with 20
mL of degassed saturated aqueous NH4Cl. The organic layer was
dried over MgSO4 and filtered, and the solvent was removed in
vacuo. The crude oil was purified by SiO2 column chromatography
(5% EtOAc in hexane) to give an off-white solid. Yield: 1.13 g
(53%). Mp = 119.3 °C. 1H NMR (500 MHz; CDCl3; 298 K): δ
7.47 (d, J = 7.4 Hz, 1H), 7.38-7.35 (m, 10H), 7.06 (t, J = 7.5 Hz,
1H), 7.00 (d, J = 7.0 Hz, 1H), 2.46 (s, 3H), 2.30 (s, 3H) ppm. 13C
NMR (126 MHz; CDCl3; 298 K): δ 137.8, 136.4 (d, JCP = 9.0 Hz),
133.2, 133.0, 128.93, 128.88, 126.7 (d, JCP = 18.4 Hz), 123.7, 122.9
(d, JCP = 27.2 Hz), 120.2, 119.7, 117.1 (d, JCP = 1.6 Hz), 16.5, 10.1
(d, JCP = 11.1 Hz) ppm. 31P NMR (202 MHz; CDCl3; 298 K): δ
-31.56 (s) ppm. HRMS (FAB): calcd for [M þ H]þ C22H21NP,
330.1412; found, 330.1417.
Synthesis of Taddol-Based IndolPhos Ligand 2c. To a solution
of di-o-tolyl(3-methyl-2-indolyl)phosphine (0.166 g, 0.48 mmol) in
THF (5 mL) was added dropwise 0.20 mL of n-BuLi (2.5 M in
hexanes) at -78 °C. The resulting yellow solution was stirred for
0.5 h at -78 °C, after which a solution of the phosphorochloridite
derived from (4R,5R)-2,2-dimethyl-R,R,R0,R0-tetraphenyl-1,3-diox-
olane-4,5-dimethanol (0.256 g, 0.48 mmol) in THF (5 mL) was
added. The reaction mixture was stirred overnight, allowing it to
warm to room temperature. The resulting pale yellow solution was
evaporated to dryness, and the resulting residue was purified by
flash SiO2 column chromatography (2% EtOAc in hexane) to give
a white solid. Yield: 0.258 g (64%). R2D2 = þ113.2 (c 1.0, CHCl3).
Mp = 127.5 °C. 1H NMR (500 MHz; CDCl3; 223 K): δ 8.81 (d,
J = 8.0 Hz, 0.4H), 8.67 (d, J = 8.8 Hz, 0.6H), 7.85 (br s, 1H), 7.69
(d, J = 8.0 Hz, 0.6H), 7.63 (m, 3.4H), 7.55 (t, J = 7.8 Hz, 1H), 7.45
(d, J = 6.9 Hz, 2H), 7.38-7.08 (m, 21H), 6.91-6.88 (m, 1H), 6.77
(dd, J= 7.6, 3.9 Hz, 1H), 6.41 (br s, 1H), 5.42 (d, J= 8.1 Hz, 0.4H),
5.28 (dd, J = 8.3, 2.1 Hz, 0.6H), 4.93 (d, J = 8.4 Hz, 0.4H), 4.91 (d,
J= 8.4 Hz, 0.6H), 2.26 (s, 1.2H), 2.24 (s, 1.8H), 2.07 (s, 3H), 1.65 (s,
3H), 1.59 (s, 1.2H), 1.50 (s, 1.8H), 0.22 (s, 3H) ppm. 13C NMR (126
MHz; CDCl3; 298 K): δ 146.2, 145.1, 140.9, 138.6, 133.1, 130.2 (d,
JCP = 5.0 Hz), 129.1, 128.4, 128.2, 127.4, 127.33, 127.31, 127.27,
123.4, 121.2, 118.9, 116.4, 112.2, 83.2 (d, JCP = 9.5 Hz), 83.2, 82.8,
58.5, 57.4, 27.8, 25.4, 21.3 (d, JCP = 20.1 Hz) ppm. 31P NMR (202
MHz; CDCl3; 223 K): δ 141.10 (d, J = 218.4 Hz, 0.4P), 139.89 (d,
136.9, 136.3, 135.0, 133.2, 132.5 (d, JCP = 18.5 Hz), 131.9 (d, JCP
=
18.4 Hz), 129.3 (d, JCP = 4.4 Hz), 128.9, 128.7, 128.5 (d, JCP = 6.1
Hz), 128.3 (d, JCP = 5.8 Hz), 128.1, 128.0, 127.7, 126.9, 126.5, 125.9
(d, JCP = 3.3 Hz), 125.2, 124.7, 123.1, 121.5, 119.1, 117.3, 111.8,
84.1, 84.0, 83.8, 83.6, 82.2 (d, JCP = 5.4 Hz), 28.1, 25.5, 21.7, 21.68,
21.65, 21.5, 10.6 ppm. 31P NMR (202 MHz; CDCl3; 298 K): δ
140.83 (d, J = 209.9 Hz, 1P), -28.73 (d, J = 210.4 Hz, 1P) ppm.
HRMS(FAB):calcdfor[Mþ H]þ C60H62NO4P2, 922.4154; found,
922.4157.
Synthesis of Taddol-Based IndolPhos Ligand 2e. To a solution
of diisopropyl(3-methyl-2-indolyl)phosphine (0.247 g, 1.00 mmol)
in THF (10 mL) was added dropwise 0.42 mL of n-BuLi (2.5 M in
hexanes) at -78 °C. The resulting yellow solution was stirred for
0.5 h at -78 °C, after which a solution of the phosphorochloridite
derived from (4R,5R)-2,2-dimethyl-R,R,R0,R0-tetrakis(3,5-dimethyl-
phenyl)-1,3-dioxolane-4,5-dimethanol (0.643 g, 1.05 mmol) in THF
(10 mL) was added. The reaction mixture was stirred overnight,
allowing it to warm to room temperature. The resulting pale yellow
solution was evaporated to dryness, and the resulting residue was
purified by flash SiO2 column chromatography (2% EtOAc in
hexane) to give a white solid. Yield: 0.580 g (68%). R2D2 = -93.6 (c
1.0, CHCl3). Mp = 133.3 °C. 1H NMR (500 MHz; CDCl3; 318 K):
δ 8.47 (d, J = 8.5 Hz, 1H), 7.57 (d, J = 7.9 Hz, 1H), 7.46 (s, 2H),
7.29 (t, J = 7.2 Hz, 1H), 7.23 (t, J = 7.4 Hz, 1H), 7.12 (s, 2H), 7.10
(s, 2H), 6.97 (s, 2H), 6.95 (s, 1H), 6.86 (s, 1H), 6.82 (s, 1H), 6.81 (s,
1H), 5.74 (dd, J = 8.0, 4.3 Hz, 1H), 4.99 (d, J = 8.2 Hz, 1H), 2.46
(s, 3H), 2.37 (br m, 1H), 2.34 (s, 6H), 2.29 (br m, 1H), 2.24 (s, 6H),
2.23 (s, 6H), 2.19 (s, 6H), 1.43 (s, 3H), 1.08-1.03 (m, 3H),
0.92-0.82 (m, 6H), 0.58-0.53 (m, 3H), 0.48 (s, 3H) ppm. 13C
NMR (126 MHz; CDCl3; 298 K): δ 145.5 (d, JCP = 2.2 Hz),
142.0, 141.6, 139.0, 137.2, 136.9, 136.8, 136.1, 129.5, 129.3, 128.9,
127.5, 126.5, 125.9, 125.2, 122.9, 121.2, 119.2, 117.6, 112.4, 110.7,
83.3, 83.2, 82.9, 27.9, 26.1, 25.5, 24.6, 21.8, 21.7, 21.6, 20.5 (d,
JCP = 19.5 Hz), 19.8 (d, JCP = 7.8 Hz), 11.3 (d, JCP = 15.8 Hz)
ppm. 31P NMR (202 MHz; CDCl3; 318 K): δ 139.89 (d, J = 61.3
Hz, 1P), -7.þ67 (d, J = 58.0 Hz, 1P) ppm. HRMS (FAB): calcd
for [M þ H] C54H66NO4P2, 854.4467; found, 854.4484.
Synthesis of Taddol-Based IndolPhos Ligand 2f. To a solution
of diphenyl(3-methyl-2-indolyl)phosphine (0.196 g, 0.62 mmol) in
THF (5 mL) was added dropwise 0.26 mL of n-BuLi (2.5 M in
hexanes) at -78 °C. The resulting yellow solution was stirred for
0.5 h at -78 °C, after which a solution of the phosphorochloridite
derived from (4R,5R)-2,2-dimethyl-R,R,R0,R0-tetrakis[3,5-bis-
(trifluoromethyl)phenyl]-1,3-dioxolane-4,5-dimethanol (0.800
g, 0.74 mmol) in THF (5 mL) was added. The reaction mixture
was stirred overnight, allowing it to warm to room temperature.
The resulting pale yellow solution was evaporated to dryness,
and the resulting residue was purified by flash SiO2 column

2776 Organometallics, Vol. 29, No. 12, 2010
Wassenaar et al.
chromatography (4% EtOAc in hexane) to give a white solid.
Yield: 0.352 g (45%). R2D2 = -60.0 (c 1.0, CHCl3). Mp = 105.4 °C.
1H NMR (500 MHz; CDCl3; 298 K): δ 8.31 (s, 2H), 8.23 (s, 1H),
7.99 (s, 2H), 7.93 (s, 1H), 7.91 (s, 2H), 7.86 (s, 1H), 7.83 (s, 3H),
7.78 (s, 1H), 7.66 (d, J = 7.8 Hz, 1H), 7.49 (t, J = 7.8 Hz, 1H),
7.40 (t, J = 7.5 Hz, 1H), 7.31 (br s, 3H), 7.27-7.19 (m, 5H), 7.12
(t, J = 7.5 Hz, 2H), 5.43 (dd, J = 8.2, 3.4 Hz, 1H), 4.75 (d,
controller. The apparatus is suited for monitoring gas-uptake
profiles during the catalytic reactions. The autoclaves were heated
to 110 °C and flushed with argon (22 bar) five times. Next, the
reactors were cooled to 25 °C and flushed again with argon (22 bar)
five times. The autoclaves were charged with the appropriate
amount of [Rh(acac)(CO)2], ligand, substrate, and internal stan-
dard in 5.00 mL of toluene under argon. The reactors were heated to
40 °C and pressurized to the desired pressure with CO/H2 (1:1), and
the pressure was kept constant during the whole reaction. The
reaction mixtures were stirred at 40 °C for 17 h, and the syngas
uptake was monitored and recorded for every reactor. After
catalysis the pressure was reduced to 2.0 bar, and samples were
taken for GC and chiral GC analysis.
High-Pressure NMR Experiments. In a typical experiment a
5 mm sapphire high-pressure NMR tube was filled with a solution
of [Rh(acac)(CO)2] (4.0 mg, 0.016 mmol), ligand (0.017 mmol), and
benzene-d6 (0.5 mL) under an argon atmosphere. Formation of the
acac complex [Rh(acac)(L)] was checked by NMR. The tube was
then purged three times with 5 bar of CO/H2 (1:1), pressurized with
5 bar of CO/H2 (1:1), and left for 16 h at 50 °C. After cooling to
room temperature the NMR spectra were recorded at the desired
temperature.
J = 8.5 Hz, 1H), 1.71 (s, 3H), 1.58 (s, 3H), 0.46 (s, 3H) ppm. 13
NMR (126 MHz; CDCl3; 298 K): δ 145.9, 145.2 (d, JCP = 2.4
Hz), 141.9, 141.6, 138.6, 134.2 (t, JCF = 6.7 Hz), 133.9 (t, JCF
C
=
5.4 Hz), 133.7, 132.9 (d, JCP = 10.7 Hz), 132.6, 132.4, 132.3 (d,
JCP = 3.7 Hz), 132.1 (d, JCP = 3.0 Hz), 132.0, 131.9, 131.8,
131.7, 131.5, 131.3, 129.2, 128.73, 128.67, 128.65, 128.60, 128.5,
127.1, 126.4, 126.3, 126.3, 126.2, 124.9, 124.20, 124.17, 124.12,
124.0, 123.5, 122.8, 122.7, 122.03, 122.00, 121.9, 121.8, 120.2,
119.9, 119.83, 119.77, 119.6, 114.6, 114.1, 82.8, 82.2 (d, JCP
=
11.1 Hz), 81.41, 81.36, 81.2, 27.2, 25.6, 10.40 ppm. 31P NMR
(202 MHz; CDCl3; 298 K): δ 138.93 (d, J = 198.9 Hz, 1P),
-26.10 (d, J = 198.8 Hz, 1P) ppm. HRMS (FAB) calcd for [M þ
H]þ C60H38F24NO4P4, 1354.1893; found, 1354.1904.
Synthesis of Taddol-Based IndolPhos Ligand 2g. To a solution
of diphenyl(3,7-dimethyl-2-indolyl)phosphine (0.310 g, 0.94 mmol)
in THF (10 mL) was added dropwise 0.40 mL of n-BuLi (2.5 M in
hexanes) at -78 °C. The resulting yellow solution was stirred for
0.5 h at -78 °C, after which a solution of the phosphorochloridite
derived from (4R,5R)-2,2-dimethyl-R,R,R0,R0-tetraphenyl-1,3-dioxo-
lane-4,5-dimethanol (0.525 g, 0.99 mmol) in THF (5 mL) was added.
The reaction mixture was stirred overnight, allowing it to warm
to room temperature. The resulting pale yellow solution was evapo-
rated to dryness, and the resulting residue was purified by flash
SiO2 column chromatography (3% EtOAc in hexane) to give a
white solid. Yield: 0.327 g (42%). R2D2 =þ94.4 (c1.0, CHCl3).Mp=
189.1 °C. 1H NMR (500 MHz; CDCl3;233K):δ7.83 (br s, 2H), 7.58
(t, J = 7.2 Hz, 2H), 7.42 (t, J = 6.7 Hz, 4H), 7.38-7.33 (m, 8H),
7.28-7.21 (m, 8H), 7.12 (t, J = 7.5 Hz, 2H), 7.05 (ddd, J = 21.3,
14.5, 7.1 Hz, 5H), 6.87 (t, J = 7.7 Hz, 2H), 5.40 (dd, J = 8.3, 2.5 Hz,
1H), 5.22 (d, J = 8.4 Hz, 1H), 1.94 (s, 1.5H), 1.92 (s, 1.5H), 1.69 (s,
3H), 1.51 (s, 3H), 0.12 (s, 3H) ppm. 13C NMR (126 MHz; CDCl3;
298 K): δ147.0 141.9, 140.7, 136.6 (t, JCP = 16.1 Hz), 132.1, 131.7 (d,
JCP = 18.3 Hz), 129.8 (d, JCP = 8.0 Hz), 129.0 (d, JCP = 6.2 Hz),
128.3, 127.7, 127.3, 121.1, 116.8, 112.1, 84.8, 84.3, 83.2, 83.1, 82.6,
27.9, 24.9, 11.5 ppm. 31P NMR (202 MHz; CDCl3; 233 K): δ 140.88
(d, J=6.7Hz, 1P), -23.08 (d, J=6.1Hz, 1P) ppm. HRMS(FAB):
calcd for [M þ H]þ C53H48NO4P2, 824.3059; found, 824.3054.
General Procedure for Asymmetric Hydroformylation. The
experiments were carried out in a stainless steel autoclave (150
mL) charged with an insert suitable for eight reaction vessels
(including Teflon mini-stirring bars) for conducting parallel reac-
tions. Styrene was filtered freshly over basic alumina to remove
possible peroxide impurities. The autoclave was charged with 1.0
μmol of[Rh(acac)(CO)2], 2.0 μmol of ligand, 1.0 mmol of substrate,
and 0.50 mmol of internal standard (decane for styrene and allyl
cyanide, heptane for vinyl acetate) in a total volume of 1.0 mL using
toluene as solvent. Before starting the catalytic reactions, the
charged autoclave was purged three times with 15 bar of syngas
(CO/H2 = 1:1) and then pressurized at 20 bar (CO/H2 = 1:1). The
reaction mixtures were stirred at 40 or 70 °C for 20 h. After catalysis
the pressure was reduced to 1.0 bar, and the conversion and b/l ratio
were checked by GC (BPX35 column, FID detector). The enantio-
meric purity was determined by chiral GC. For styrene: Supelco
BETA DEX 225 column; initial temperature = 100 °C for 5 min,
then 4 °C/minto160°C; tR(S) = 11.27 min and tR(R) = 11.45 min.
For vinyl acetate: Supelco BETA DEX 225 column; initial tem-
perature = 100 °C for 5 min, then 4 °C/min to 160 °C; tR(S) = 6.14
min and tR(R) = 7.85 min. For allyl cyanide: Astec Chiraldex A-TA
column; initial temperature = 90 °C for 7 min, then 5 °C/min to
180 °C; tR(S) = 16.29 min and tR(R) = 16.75 min.
High-Pressure in Situ IR Spectroscopy. In a typical experiment
a 50 mL HP IR autoclave was filled with a solution of ligand (0.030
mmol) in cyclohexane (15 mL). The autoclave was purged three
times with 15 bar of CO/H2 (1:1), pressurized with 20 bar of CO/H2
(1:1), and heated at 40 °C under stirring, and a background spec-
trum was recorded. A solution of [Rh(acac)(CO)2] (3.9 mg, 0.015
mmol) in cyclohexane (1 mL), previously charged into the reservoir
of the autoclave, was added to the reaction mixture by overpressure.
After complete conversion to the hydridobiscarbonyl rhodium
species a solution of vinyl acetate (0.28 mL, 3.0 mmol) in cyclohex-
ane (1 mL), previously charged into the reservoir of the autoclave,
was added to the reaction mixture by overpressure. The IR spectra
were recorded at 40 °C.
DFT Calculations. The geometry optimizations were carried
out with the Turbomole program25 coupled to the PQS Baker
optimizer.26 The geometries were optimized as minima at the ri-
DFT BP8627 level using the SV(P) basis set28 on all atoms. We
applied the BOpt script package developed by Prof. Peter H. M.
Budzelaar to facilitate all data evaluation processes.29 The ob-
tained stationary points (minima) were characterized by vibra-
tional analysis (numerical frequencies); ZPE and gas phase
thermal corrections (entropy and enthalpy, 298 K, 1 bar) from
these analyses were calculated according to standard formulas of
statisticalthermodynamics.Thethusobtained(free) energies(kcal
mol-1) are reported in Table 6.
Acknowledgment. This work was supported by NRSC-
C and the European Union (RTN Revcat MRTN-CT-
2006-035866). We thank V. Bocokic for assistance with
the high-pressure IR experiments and Dr. J. I. van der
Vlugt for fruitful discussions.
Supporting Information Available: High-pressure 1H and 31
P
NMR spectra for the hydridobiscarbonyl rhodium complexes, and
NMR spectra for all new ligands and deuterioformylation pro-
ducts. This material is available free of charge via the Internet at
http://pubs.acs.org.
(25) (a) Ahlrichs, R. Turbomole Version 5; Theoretical Chemistry
€
Group, University of Karlsruhe, 2002. (b) Schafer, A.; Huber, C.;
Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829–5835.
(26) (a) PQS version 2.4; Parallel Quantum Solutions: Fayettevile, AR,
2001 (the Baker optimizer is available separately from PQS upon request). (b)
Baker, J. J. Comput. Chem. 1986, 7, 385–395.
(27) (a) Becke, A. D. Phys. Rev. A: Gen. Phys. 1988, 38, 3098–3100. (b)
Perdew, J. P. Phys. Rev. B 1986, 33, 8822–8824.
Kinetic Gas-Uptake Measurements. The experiments were
carried out in an AMTEC SPR16 consisting of 16 parallel reactors
equipped with temperature and pressure sensors and a mass flow
€
(28) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97,
2571–2577.
(29) Budzelaar, P. H. M. J. Comput. Chem. 2007, 28, 2226–2236.